Forums : The Ohcount Project
Dear Open Hub Users,
We’re excited to announce that we will be moving the Open Hub Forum to
https://community.blackduck.com/s/black-duck-open-hub.
Beginning immediately, users can head over,
register,
get technical help and discuss issue pertinent to the Open Hub. Registered users can also subscribe to Open Hub announcements here.
On May 1, 2020, we will be freezing https://www.openhub.net/forums and users will not be able to create new discussions. If you have any questions and concerns, please email us at
info@openhub.net
Stats wrong for Thousand Parsec
Hello,
I would like to report that ohcount is not correctly detecting the code formats for our repositories. It is saying that our code is 27% shell script when it is much closer to 2%.
You can see the stats at http://www.ohloh.net/projects/thousand-parsec/analyses/latest
and you can find out code at
http://git.thousandparsec.net
I'm happy to try and help you figure out what is going wrong.
Tim 'Mithro' Ansell
We also run sloccount on our own code. You can also see a copy of what sloccount knows about at the following page -
http://www.thousandparsec.net/tp/dev/sloc.php
Clearly we are mostly C++, followed closely by Python. We also have a bit of Java and pure C code too.
Hi Tim,
Our sloccount expert will be back Monday. If you can wait until then, we'll be able to explain/investigate more efficiently.
-Andy
PS Make, config, and other build related files might be contributing to the high shell script %.
This is actually a long running problem. We where going to try and figure out what was wrong last time I noticed this, but I think everyone got distracted.
We only have make, config and other build related files for tpserver-cpp, so I can't see them dwarfing all our other code.
Hi mithro,
It took me a long time to check out all 41 repositories, but the mystery is pretty simply solved.
Sloccount ignores configure files, while Ohcount counts them as a shell scripts.
Since tpserver-cpp is redundantly included in the Ohloh report four separate times, once for four different branches, Ohloh is counting the code in this directory four times over, which really adds up to a lot of configure files.
In general, the list of code at http://www.thousandparsec.net/tp/dev/sloc.php only lists about half of the code listed on Ohloh, so there is a great deal of code that Ohloh is counting (or overcounting in the case of multiple branches) that the sloc.php page does not report. This makes direct comparisons basically impossible.
Ohcount does have special language categories for autoconf and automake, and it categorizes files like configure.in as autoconf files instead of shell scripts. I'm not sure why we count the configure file itself as shell script -- it seems to me that we should count it as an autoconf file.
Our make/automake/autoconf support was contributed by ciaranm -- Ciaran, any chance you are looking at this thread and can offer an opinion on this?
Unless someone objects, I'm sort of inclined to remove configure from the shell script category and label it as autoconf.
You can download our line counting tool Ohcount from http://labs.ohloh.net to answer these kinds of questions -- it can give detailed counts, and identify files by language -- you can pretty rapidly find out where the shell script content is coming from.
Thanks,
Robin
